1. Field of the Invention
The present invention relates to an information processing apparatus and information processing method for administering an evaluation system (a set of units for evaluation) which provides evaluation including classification, providing comments or the like to a data object, for example, a document that can be electronically accessed.
2. Discussion of the Related Art
The environment for utilizing electronic information has been recently developed by introducing word processors, various types of filing systems, database systems, network systems and so on to offices. The patents, papers, many kinds of references or the like in all over the world are now stored in the on-line database, patent information is stored in CD-ROM, and moreover, the Internet has been widespread, whereby the amount of electronic information available in the offices is explosively increased. Accompanying thereto, importance of sharing information in organizations in the offices has been recognized.
However, most of the electronic information utilized in the organizations is an electronic replacement of information conventionally represented in paper documents. With only such replacement, it is difficult to say that the advantages generated by the use of the electronic information are not fully obtained. For appropriately acquiring the necessary information, evaluation including classification, providing comments and so forth should be provided to the information according to unique meanings in the organization.
Conventionally, several methods of mutual offering of information which has a unique meaning were proposed for accelerating the information sharing. One of them is to share the comments for the information, which is disclosed by Japanese Patent Application Laid-Open Nos. 6-274493 (1994) and 7-311764 (1995). These inventions have a purpose to support the collaborative writing, which administer comments given to the document by each member and make it possible to retrieve and display the comments. Each member gives arbitrary comments, and consequently, there is no problem if information is shared in a group of the small number of people, but in the case where information is shared in a large organization, or a large amount of pieces of information are collected in a long period, there occurs a deficiency that it is difficult to find from what viewpoint the comments are given to each piece of information. For example, if only the comment "Important" has been given to a piece of information, it is impossible to know for what purpose the information is "important".
As a tool of offering subjective ratings about an article to one another, for example, GroupLens disclosed by "GroupLens: An Open Architecture for Collaborative Filtering of Netnews", Resnick et al., CSCW94, pp. 175-186, 1994 is known. This makes it possible for users to freely provide 5-grade ratings to every article in the news on an electronic bulletin board. When a user refers to an article, the user can determine the degree of importance with reference to ratings given by other users.
However, there is a problem that GroupLens is poor in expression because the basic rating system such as classification is fixed. Even if the rating system could be renewed, it would be necessary to unify the will of the users for halfway changing of the rating system. Therefore, it is difficult to create a new rating system.
There is also a method of constructing classification categories for uniquely providing a meaning to each of pieces of information and classifies the pieces of information using the classification categories.
Inventions disclosed by Japanese Patent Applications Laid-Open Nos. 5-63834 (1993), 7-28834 (1995) and 7-110809 (1995) store the classification categories and carry out the proper choices or re-classification of information such as documents by selecting the items corresponding to the object information from the categories. An invention disclosed by Japanese Patent Application Laid-Open No. 7-84858 (1995) also administers the classification categories, and in particular administers files from plural classification viewpoints.
However, a system for providing meanings to information in an organization may be changed by various external elements such as reorganization, change of research direction, the trend of technology and reconstruction of the technological map by development of new technologies. Therefore, if the classification is continuously carried out for a long period, the change of the classification categories is required in the course of carrying out the classification. This causes a problem that it is hard to recognize by at what time of classification categories each piece of information was classified.
For example, between classification systems having categories "large" and "small", and categories "large", "medium" and "small", classification into the same category "small" has different meanings. Accordingly, if the category "medium" is added, it is necessary to reconsider whether each of all pieces of information previously classified into "large" or "small" belongs to the category "medium" or not.
In conventional methods of category administration, the reconsideration of the classification was impossible because the categories before updating were lost. If a user carries out such reconsideration immediately before the updating, it costs a great deal. Besides, in the case where the categories are shared, if a person changes the categories from his/her own viewpoint, it results in that the categories are changed without the approval of other people using the categories. Therefore in the conventional methods, it has been required that the classification categories have been continuously used as they were fixed, or that the categories have been changed on the consensus of the people after a thorough discussion.
Another method is to store the categories in the past and reproduce one of them. An invention disclosed by Japanese Patent Application Laid-Open No. 4-47366 (1992) stores the history of data structures in the past and reproduces a data structure at a certain point of time; thereby it is possible to perform information retrieval based on the reproduced data structure.
However, the purpose of the invention is to restore the data structure at a certain point of time based on the operations in the past. The plural data structures are not under version administration by names, for example, and relations among the plural data structures in the past are not stored. Therefore, according to this invention, a data structure after updating cannot be shown against a specific data structure in the past to support re-providing of evaluation, and besides, it is impossible to select necessary items from the classification categories used in the past to create new classification categories.
The evaluation categories used for evaluation such as classification or providing comments are not the given ones, but they are required to be changed in the course of collecting information. Accordingly, to support information evaluation in the organization in a long period of time, it is necessary to resolve the following problems.
(1) Though it is possible to make a lot of variations of the evaluation categories in accordance with the change of the environment or customization by each member, it is difficult to select or construct the categories which include items sufficient to provide evaluation.
(2) When the evaluation categories are referred to, it is impossible to recognize whether the categories are the latest ones or not. If the categories are not the latest, it is required to provide evaluation again, but it costs very much.